BackgroundVery early in my amateur career, around 1962, I came across an American VHF bulletin In which I found an advertisement for "spiral ray" antennas, this was a type of yagi in which the elements were not parallel - the elements were arranged to gradually twist from -45 degrees to +45 degrees in such a way that the antenna produced a more or less circular polarisation. In this advertisement it was stated that this arrangement was particularly well suited for aurora. Since my interest already was DX via aurora on 144MHz, I immediately started to investigate the polarisation of aurora signals by standing on the balcony holding a 10 element antenna in my hands. The result I got was very clear: The polarisation plane is conserved. There was never any signal with the antenna vertically polarised, and consequently a circularly polarised array would be a certain 3dB loss.
Now I know that the result from 1962 was wrong. The reason is that at that time all stations I could check were more or less north or south of my QTH. About 10 years later SM5AGM told me that he, during a major aurora had observed that a station in Moscow was vertically polarised. Since that time I have been experimenting with polarisation on aurora, gradually more and more carefully.
Polarisation of Aurora signalsThere are a few statements that can definitely be made, and knowing the facts help a great deal to get the pleasure of working many and distant stations via aurora. Parts of these statements I think are well known, but other parts I believe are more or less unknown to the amateur community. All my observations are made from grid JO89, and there is of course a possibility that the aurora propagation is different at other geomagnetic latitudes.
1. In a north to south path, the polarisation is always conserved, regardless of the distance and intensity of the aurora. The easiest way to work long distance to the south is to elevate the antenna during strong auroras. The optimum elevation may be as high as 25 degrees during very strong events.
2. Regardless of the communication path the polarisation is always conserved during minor auroras. An EME capacity station 1500km east or west will always be horizontal, if the signal level is below S7 or so in the optimum polarisation plane.
3. During strong auroras the polarisation plane is twisted, and the polarisation offset depends on aurora intensity and the east to west component off the transmission path. (The polarisation always stays linear, but the best antenna direction may differ with different polarisations)
The statement 3 above is the important one: If you point a vertically polarised array somewhere around 300 or 60 degrees in azimuth, the stations within about 500km disappear, while the stations at 1500km become strong, provided the aurora event is a good one. (From my QTH there are very few stations at longer distances east or west).
Circular is fine for auroraHaving a station designed for EME with polarisation control, the loss of 3dB during transmit is usually no problem. During an aurora it is possible to work distant stations with moderate power, and they will not notice a 3dB loss of signal. But when linear polarisation is used, it is quite common to loose 10 or 20 dB due to faraday rotation - and then EME power is not enough. When calling CQ for eastern or western stations, I usually use circular polarisation because that provides more answers. For receive I do not use circular - I use linear, and tune the polarisation plane for optimum S/N - often there is a pile up, and changing the polarisation plane is one way to separate different stations.
Selecting the best linear polarisation for auroraSince my system only allows horizontal, vertical and circular polarisation, my empirical knowledge is limited. I guess the rule for EME is valid with spatial polarisation offset = 0. The practical conclusion is: always transmit with the same polarisation as the one giving best received signal if you can choose only between horizontal and vertical.
Does the aurora reflection depend on polarisationIt has been suggested that the aurora reflection may be very much stronger with vertical than with horizontal polarisation. (QST dec 1994)
Since I am using QSK, the station operates like a narrow band radar. Aurora and other Echoes With Narrow Band Radar In the old days, from about 1975 varicaps were used for the switch, but now they are replaced with PIN diodes. High power antenna relay using PIN diodes
I have tried many times to compare the signal level with different polarisations, and there is a definite conclusion: There is no difference between vertical and horizontal, but circular is no good!
The reason for circular polarisation to not work well in radar mode is that the same circular polarisation is used for transmit and receive. A large attenuation is quite natural - the circular polarisation is expected to reverse its rotation on reflection.
Of course radar experiments will only tell what happens when the wave is reflected by 180 degrees. For long distance communication, the angle between incoming and outgoing wave may be about 90 degrees - and my experiments give no indication on what would happen then.